CONTENT CREATION

Nightly sweep to caption Airtable assets missing alt-text

Runs on a schedule, finds Airtable asset records that have no alt-text, generates one with HuggingFace, and posts a Slack summary of how many were filled and which need attention.

CategoryContent Creation
Enginesim
Difficultyintermediate
Triggerschedule
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNightly schedule fires
  • ActionQuery Airtable for assets missing alt-textAirtableAirtable
  • ActionGenerate captions with HuggingFace modelHugging FaceHugging Face
  • LogicBranch on caption confidence (Done vs Needs Review)
  • ActionWrite captions and status back to AirtableAirtableAirtable
  • OutputPost cleanup summary to SlackSlack

What it does

On a nightly schedule, this workflow scans your Airtable asset library for records where the alt-text field is empty, generates captions for them with a HuggingFace model, and reports the cleanup results to Slack.

When to use it

Use it when you already have a large existing media library with gaps, not just new uploads. It steadily chips away at historical assets that were never described, so accessibility coverage trends toward 100 percent without a manual audit.

How it works

  1. 1A scheduled trigger fires once per night.
  2. 2Airtable is queried for asset rows where the alt-text field is blank.
  3. 3For each record, the image URL stored on the row is sent to a HuggingFace image-to-text model.
  4. 4A branch checks model confidence: high-confidence captions are written straight back to Airtable as Done; low-confidence ones are written with a Needs Review flag.
  5. 5A Slack message summarizes counts filled, flagged, and remaining for the team.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect AirtableBases, tables, views, automations.
  2. 2
    Connect Hugging FaceModels, datasets, spaces — the open-source hub.
  3. 3
    Connect SlackChannels, DMs, threads, mentions.
  4. 4
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  5. 5
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  6. 6
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.